Encoding an image includes determining respective costs of coding a symbol using available coding types. A first coding type indicates that a value of the symbol is to be decoded using a same number of bits, and a second coding type indicates that the value of the symbol is to be decoded using a range. An optimal coding type of the available coding types is selected, which corresponds to a smallest cost of the respective costs. A first indicator of the optimal coding type and a first symbol value of the symbol using the optimal coding type are encoded in a compressed bitstream. Decoding an image includes decoding, from a header of a compressed bitstream, respective coding types of symbols encoded in the compressed bitstream and decoding, from the compressed bitstream, respective values of the symbols according to the respective coding types decoded from the header.

Concepts are presented which achieve a more efficient coding of coefficients of a transform block by use of dependent quantization and context adaptive entropy coding or achieve a coding of coefficients of a transform block in a manner which allows a more efficient coding even if a usage of dependent quantization is combined with the usage of context adaptive entropy coding.

An encoder includes circuitry and memory coupled to the circuitry. The circuitry, in operation, for each coefficient of a plurality of coefficients included in a block, determines a base level relating to Context-Based Adaptive Binary Arithmetic Coding (CABAC) for the coefficient, and encodes an absolute value of the coefficient. In determining the base level, when one or more flags are used in encoding the absolute value of the coefficient, the base level is determined to be a first value, and when one or more flags are not used in the encoding, the base level is determined to be a second value that is smaller than the first value. In encoding the absolute value of the coefficient, when one or more flags are not used, a rice parameter is determined based on the base level which is equal to the second value, and the coefficient is binarized using the rice parameter.

An apparatus and a method for filtering reconstructed images, in particular, video images, with adaptive multiplicative filters. The efficiency of the filtering operation is increased by restricting the allowable values of the filter coefficients to those that have only a limited number of “ones” in the binary representation.

According to the disclosure of the present document, the resources necessary for an LMCS process may be reduced by limiting the range that an LMCS APS occupies in all APSs. Thus, the degree of complexity in LMCS may be reduced, and video/image coding efficiency may be increased.

Embodiments of this disclosure provide an image coding method and apparatus and an image compression system. The image coding apparatus includes a memory and a processor. The processor is configured to perform feature extraction on an input image to obtain feature maps of N channels; assign a weight to a feature map of each channel among the N channels; perform down-dimension processing on weighted feature maps processed in association with the N channels, to obtain feature maps of M channels and output the feature maps of M channels, M being smaller than N. Hence, by multiplying different feature maps by a weight to obtain corresponding importance and then performing down-dimension processing on the feature maps processed according to the weighting, time for decoding may be reduced.

An image decoding method according to the present document comprises a step for performing inverse first-order transform and inverse non-separable transform on a residual sample. The inverse non-separable transform is performed on the basis of a transform index indicating a predetermined transform kernel matrix, the inverse first-order transform is performed on the basis of a multiple transform selection (MTS) index indicating MTS for a horizontal transform kernel and a vertical transform kernel, and a syntax element bin string for the transform index is derived on the basis of first context information when a tree type for a split structure of a target block is not a single tree type and is derived on the basis of second context information when the tree type is the single tree type.

An image quantization parameter encoder includes: a prediction unit for generating a predicted quantization parameter from a past reconstructed quantization parameter; a computing unit for generating a delta quantization parameter from a quantization parameter and the predicted quantization parameter; and a quantization parameter encoding unit for binary-arithmetic-encoding a first bin indicating whether the delta quantization parameter is significant; and when the delta quantization parameter is determined to be significant, other bins, which are encoded subsequent to the first bin, indicating an absolute value of the delta quantization parameter, and a sign bin, which is encoded subsequent to the other bins, indicating whether the delta quantization parameter is positive or negative, wherein the binary-arithmetic-encoding is performed using contexts for the first bin and the other bins, and using no context for the sign bin, and wherein the binary-arithmetic-encoding is performed using a first context for the first bin and using a second context for the other bins.

A device for coding video data is configured to code a motion vector difference (MVD) value using a codeword prefix and a codeword suffix. Based on an absolute value, minus 2, of the MVD value being 131070 or greater, the codeword prefix of the MVD value is equal to 1111 1111 1111 1111 and the codeword suffix is coded using the fixed length code.

There is provided an video encoding/decoding method and apparatus. The video decoding method comprises acquiring a bitstream including a predetermined context element, performing at least one of a context model determination, a probability update, and a probability interval determination on the predetermined syntax element, and arithmetically decoding the predetermined syntax element on the basis of a result of the performance.